M. Pereda Vivo, C. Paroissin
In many areas (like toxicology, chemistry, and more generally in environmental science), when dealing with concentration measurements with an analytical method, one will observe an exact measurement only if it is larger than a certain threshold, called limit of detection; otherwise, one has only the information that the concentration lies between zero and this limit. Such a situation is called left-censoring. In addition, in some situations there could be the absence of the substance under consideration, we talk about data with zero excess. In this work, we develop a semi-parametric mixture model to analyse data subject to left censoring with inflation of zeros. For the continuous positive part, we consider a semi-parametric proportional reversed hazard rate model. For the zero-outcome value, we consider a parametric regression model. We estimate the parameters in the mixture model using the Expectation-Maximization algorithm. Finally, we have performed a simulation study.
Keywords: Limit of Detection, Left-censoring, Zero-inflated data
Scheduled
Ramiro Melendreras Award III
November 7, 2023 4:50 PM
CC4: Room 2